1. Home
  2. Electronics Design
  3. Ping: Vbe Thompson et al

Ping: Vbe Thompson et al

So if you peruse the application note somewhere for the crap garbage LM324, it shows an op amp current source with multiple outputs. The first output is the standard thing with neg feedback from a pnp emitter resistor to the inverting input, with the reference voltage on the non inverting.

Then to get the multiple outputs, they just hang more pnps with emitter resistors off the base of the first one which is connected to the opamp output.

That's OK I guess, but the problem is in simulation that only the first output is somewhat insulated against variations over temperature, due to the feedback from the opamp. Since the other mirrors are not in the loop, they're going to be all over the place.

So...I guess you could have an opamp section per, and feed them all from the same stable reference. But that gets annoying if you need, say, 24 outputs.

Has some analog wonk come up with a more economical solution somewhere along the line?

--


----Android NewsGroup Reader---- 
http://usenet.sinaapp.com/
Reply to
bitrex
Loading thread data ...

On Friday, January 1, 2016 at 8:24:08 AM UTC-8, bitrex wrote: ...

....

Provided the temperature of the PNPs are close there should not be much variation between outputs as they will have similar Vbe changes.

Agreed not as good as being in the feedback loop. Early effect is also not compensated for.

kevin

Reply to
kevin93

**If** the voltage across the emitter resistors is high in proportion to the delta-Vbe then it works pretty well. For example, if the

difference due to delta-Vbe is only 1%.

There is no other way, AFAIK. Of course with low dissipation, SMT parts or monolithic parts the temperature delta should be minimal and, n the latter case, Vbes will be well matched. You could try, for example, SOT-89 parts (big collector tab) with a ground plane under them for thermal coupling.

You're not going to get super high precision with this circuit anyway- there's a fraction of a percent difference in current due to beta variations and there are going to be Vbe differences between different transistors anyhow. Plus the offset voltage of a 324 is in the same order of magnitude as the Vbe variations (mV). Plus there is the Early voltage effect on the open-loop sinks.

--sp

--
Best regards,  
Spehro Pefhany 
Amazon link for AoE 3rd Edition:            http://tinyurl.com/ntrpwu8 
Microchip link for 2015 Masters in Phoenix: http://tinyurl.com/l7g2k48
Reply to
Spehro Pefhany

LM324 has limited common-mode range near plosive rail... make sure your emitter resistor drops are >= 2V

With adequately-sized emitter resistors variation should be primarily Early effect.

Post your .asc file for further assistance.

A custom chip >:-} ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.
Reply to
Jim Thompson

Use a small micro, with 24 pwm outputs (can be sw pwm)

Control the bjts with rc filtered pwm

Close the feedback loop with the micro

Reply to
Klaus Kragelund

Wow, a Jeopardy question:

Answer:

The Question:

"How could you make the LM324 look like a good op amp?"

;)

Cheers

Phil Hobbs

Reply to
Phil Hobbs

Alternate answer: "Compare it to a BJT."

I like the LM324! It's the first jellybean rail-to-NOT rail input opamp ;-). And John loves them too (ducks!).

(Waaay back I used some National JFET amps for sensing slightly over the positive rail...can't remember which ones.)

Cheers, James Arthur

Reply to
dagmargoodboat

bitrex schrieb:

You didn't tell us how much current the sources are to supply. So what about using arrays of pnps such as CA3096? Shouldn't their data be sufficiently similar?

HTH

Reinhard

Reply to
Reinhard Zwirner

Like this ??...

...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.
Reply to
Jim Thompson

In the late '70's, early '80's I used LM324's by the bucket-load... but banned Motorola's version for excessive cross-over dead-band... really screwed active filters. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.
Reply to
Jim Thompson

It's probably a multiple-collector transistor, which you can do inside an IC. TC is low and current matching (even ratio current matching) can be pretty good.

A problem in the original LM324 is if one of the current outputs swings to the rail, it can collapse the others. So opamp sections can interact in a not-nice way.

An opamp per source is the best way to get precision. Something like this might be good enough:

formatting link

How much current and voltage do you need? How accurate? Which direction?

It's sometimes fun to use an opamp power pin as a current source.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics
Reply to
John Larkin

An LM324 shows visible crossover distortion at 60 Hz!

And sections can interact. And all 4 amps can go bonkers if one input is pulled slightly negative.

DC specs are OK. With care, it can be used in some applications. The last time I designed one into a product was 11 years ago.

The LF353/347 sort of jfet amps are cheap and pretty well behaved and have a typ (not guaranteed) common-mode voltage that hits V+. They make decent comparators, which the 324 doesn't.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics
Reply to
John Larkin

[snip]

CAUTION: Saturating any single current source output will causes errors in the other outputs. ...Jim Thompson

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

formatting link
| 1962 | I love to cook with wine. Sometimes I even put it in the food.

Reply to
Jim Thompson

So, if you want a fixed amount of current, you can swamp the Vbe variation with I*R_emitter voltage drop. If one sets R_emitter to zero, a 1% current difference results from 3C temperature difference, and the 'correction' to Vbe for a 1% change is (at room temperature) about

6 mV. If you choose R_emitter so that I*R_emitter is much greater than that, it dominates.

Not insulated; nulled, unless you think the opamp gain is too low.

So, leave the opamp out entirely, do it all with the emitter resistor. Base of PNP to +12V, emitters to resistors to your new +24V power supply. With 1k ohm resistors, that gives you 11.4 mA in each leg. Many degrees C will eventually cause Vbe differences and one of your sources might go to 11.5 mA.

If 1% variation is troublesome, you need to shell out for 0.1% resistors and a snazzy regulator for the 200V power supply.

Reply to
whit3rd

Crush them both?

Reply to
krw

Yes. I fixed most of that with a pull-down to keep the output stage Class-A.

I believe you, but never saw it.

When I went to Digikey looking for R-R opamps the other day, about 3/4 of them were LM324 or kin. Lots of low-power variants.

The LM339's fun too. You can make anything with an LM339, maybe even a 555. :-)

Yep. They've got depletion FET input stages, and current draw wasn't bad for its day. I'd figured them long obsolete. Haven't looked.

Cheers, James Arthur

Reply to
dagmargoodboat

More succinct answer: "With a hammer."

Reply to
dagmargoodboat

Early op-amp chips like those said in the specs that you had to run them as Class A amps for audio. 0.1 or more volts of slop between push and pull was common.

--
I will not see posts from astraweb, theremailer, dizum, or google 
because they host Usenet flooders.
Reply to
Kevin McMurtrie

Oh, it wasn't 'slop', it was a deliberate dead-band; that way, they were power-stingy when the load was light. For battery operation, it's a good thing. That LM324 was made long after the uA741, and it was made for the low-performance market. That's a big market.

Reply to
whit3rd

Yeah but it's so cheap! (and the LM358 similarly)

It surprises me that there isn't a more modern op-amp that even comes close to being as cheap, regardless of specs.

Reply to
Chris Jones

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.